A Novel Intronic Single Nucleotide Polymorphism in the ... - Genetics

Download PDF
6 downloads 0 Views 1MB Size Report
Comment
*Department of Zoology, Ohio Wesleyan University, Delaware, Ohio 43015, ... 1Corresponding author: Ohio Wesleyan University, Schimmel/Conrades Science ...
INVESTIGATION

A Novel Intronic Single Nucleotide Polymorphism in the Myosin heavy polypeptide 4 Gene Is Responsible for the Mini-Muscle Phenotype Characterized by Major Reduction in Hind-Limb Muscle Mass in Mice Scott A. Kelly,*,1 Timothy A. Bell,† Sara R. Selitsky,† Ryan J. Buus,† Kunjie Hua,† George M. Weinstock,‡ Theodore Garland Jr.,§ Fernando Pardo-Manuel de Villena,† and Daniel Pomp†

*Department of Zoology, Ohio Wesleyan University, Delaware, Ohio 43015, †Department of Genetics and Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, North Carolina 27599, ‡The Genome Institute, Washington University School of Medicine, St. Louis, Missouri 63108, and §Department of Biology, University of California, Riverside, California 92521

ABSTRACT Replicated artificial selection for high levels of voluntary wheel running in an outbred strain of mice favored an autosomal recessive allele whose primary phenotypic effect is a 50% reduction in hind-limb muscle mass. Within the High Runner (HR) lines of mice, the numerous pleiotropic effects (e.g., larger hearts, reduced total body mass and fat mass, longer hind-limb bones) of this hypothesized adaptive allele include functional characteristics that facilitate high levels of voluntary wheel running (e.g., doubling of mass-specific muscle aerobic capacity, increased fatigue resistance of isolated muscles, longer hind-limb bones). Previously, we created a backcross population suitable for mapping the responsible locus. We phenotypically characterized the population and mapped the Minimsc locus to a 2.6-Mb interval on MMU11, a region containing 100 known or predicted genes. Here, we present a novel strategy to identify the genetic variant causing the mini-muscle phenotype. Using high-density genotyping and whole-genome sequencing of key backcross individuals and HR mice with and without the mini-muscle mutation, from both recent and historical generations of the HR lines, we show that a SNP representing a C-to-T transition located in a 709-bp intron between exons 11 and 12 of the Myosin heavy polypeptide 4 (Myh4) skeletal muscle gene (position 67,244,850 on MMU11; assembly, December 2011, GRCm38/mm10; ENSMUSG00000057003) is responsible for the mini-muscle phenotype, Myh4Minimsc. Using next-generation sequencing, our approach can be extended to identify causative mutations arising in mouse inbred lines and thus offers a great avenue to overcome one of the most challenging steps in quantitative genetics.

I

N 1993, replicated artificial selection for high voluntary wheel-running behavior was implemented, beginning from a base population of 224 outbred Hsd:ICR mice (Swallow et al. 1998). The goal was to elucidate the underpinnings of voluntary exercise at the levels of physiology, neurobiology, and motivation and eventually to identify genetic factors underlying these intermediate phenotypes. The selection criterion Copyright © 2013 by the Genetics Society of America doi: 10.1534/genetics.113.154476 Manuscript received July 10, 2013; accepted for publication September 10, 2013; published Early Online September 20, 2013. Supporting Information is available online at http://www.genetics.org/lookup/suppl/ doi:10.1534/genetics.113.154476/-/DC1. 1 Corresponding author: Ohio Wesleyan University, Schimmel/Conrades Science Center #346, 61 S. Sandusky St., Delaware, OH 43015. E-mail: [email protected]

was the total revolutions run during days 5 and 6 of a 6-day period of wheel access when mice were young adults (6–8 weeks of age). After 10 generations of selection, mice from the four replicate High Runner (HR) lines were running .70% more revolutions per day (on days 5 and 6) as compared with four nonselected control (C) lines (Swallow et al. 1998). Wheel running continued to increase in the HR lines until generation 16–28 (depending on line and sex) and remained approximately at those levels through a subsequent 50 generations of selective breeding (Middleton et al. 2008; Swallow et al. 2009; Kolb et al. 2010; Careau et al. 2012, 2013). Dissections to monitor potential correlated selection responses in organ masses led to the discovery of a mini-muscle

Genetics, Vol. 195, 1385–1395 December 2013

1385

phenotype that is caused by an autosomal recessive mutation (Garland et al. 2002; Hannon et al. 2008). The most notable feature of this phenotype is an 50% reduction in mass of the triceps surae muscle complex (gastrocnemius, plantaris, soleus) and in the mass of the entire hind-limb musculature (Houle-Leroy et al. 2003; Swallow et al. 2005). Hence, the allele qualifies as a “gene of major effect” (e.g., Agrawal et al. 2001; Hannon et al. 2008). The mini-muscle phenotype was observed in two of the four HR lines and in one of the four C lines (Table 1). Initially, the phenotype remained at low frequency in the C line and eventually disappeared, but increased dramatically in frequency in both HR lines, eventually going to fixation in one of the two HR lines by approximately generation 36 (Syme et al. 2005). Evidence for the adaptive significance of the allele was obtained by fitting multiple-generation data to hierarchical models that included effects of selection and/ or random genetic drift (Garland et al. 2002). The minimuscle allele was estimated to have been present at a frequency of 7% in the base population, and analyses indicated that strong selection favored the allele in the HR but not in the C lines. It was hypothesized that the mini-muscle phenotype has functional characteristics that facilitate high levels of wheel running (Garland et al. 2002), such as a doubled mass-specific aerobic capacity (Houle-Leroy et al. 2003) and increased fatigue resistance in at least some hind-limb muscles (Syme et al. 2005). Alternatively, the underlying allele might have pleiotropic effects that act on nonmuscle tissues and organs but still facilitate endurance running (e.g., reduced total body mass and fat mass; increased relative heart, liver, and kidney mass; longer hind-limb bones) (Garland et al. 2002; Kelly et al. 2006; Hannon et al. 2008; Kolb et al. 2010). However, at generation 22, wheel running of affected individuals did not differ statistically from those with normal-sized muscles (Garland et al. 2002), and the magnitude of response to selection was not systematically higher in the two lines that had the mini-muscle phenotype as compared with the two that did not, indicating that multiple genetic “solutions” are possible in response to selection for high levels of voluntary exercise (Garland et al. 2011a; Careau et al. 2013). To date, obvious deleterious effects of the mini-muscle phenotype have not been identified (e.g., see Girard et al. 2002), which is unusual for a gene of major effect (e.g., Carriére et al. 1994; Rivero et al. 2011). Previously, we used a backcross population between HR (line 3) and the C57BL/6J inbred strain (Hannon et al. 2008) to map the causative locus known as Minimsc (MGI:44204080) to a 2.6-Mb interval on MMU11, a region containing 100 known or predicted genes, many of which have known roles in muscle development and/or function (Hartmann et al. 2008). Here, we present a novel strategy to identify the genetic variant(s) causing the mini-muscle phenotype. Using high-density genotyping in a small but carefully selected sample of 10 mice, we first identified two maximally informative backcross samples for whole-genome sequencing, and subsequently we used the sequence data to identify

1386

S. A. Kelly et al.

Table 1 Frequency of the mini-muscle phenotype in four replicate lines selectively bred for high voluntary wheel running (High Runner, HR) and four nonselected lines that serve as their controls (Control, C) in the breeding program Line

Linetype

1 2 4 5

Control Control Control Control

3

High runner

6

High runner

7 8

High runner High runner

Mini-muscle status Never observed Never observed Never observed Observed at least through generation 22 Eventually became fixed for mini phenotype Still polymorphic for mini phenotype at generation 66 Never observed Never observed

two putative causal SNPs. Using targeted genotyping of these two SNPs in large cohorts of HR mice with and without the mini-muscle mutation, from both recent and historical generations of the HR lines, we show a SNP representing a C-to-T transition located in a 708-bp intron between exons 11 and 12 of the Myosin heavy polypeptide 4 (Myh4) skeletal muscle gene (position 67,244,850 on MMU11; assembly, December 2011, GRCm38/mm10; ENSMUSG00000057003) is responsible for the mini-muscle phenotype. The symbol of this allele is Myh4Minimsc.

Materials and Methods High-density genotyping using the mouse diversity array

To confirm and refine the Minimsc candidate interval reported previously (Hartmann et al. 2008), we genotyped 10 mice with the high-density mouse diversity array (MDA) (Yang et al. 2009). Four of these mice were selected from the backcross (BC) (Hannon et al. 2008) and used to map Minimsc, while the other six mice were selected from the four replicate HR lines (Figure 1). Samples were evenly distributed between phenotypically normal animals (BC-4197, HR7-61023, HR7-61148, HR8-61095, and HR8-61102) and animals exhibiting the mini-muscle phenotype (BC-4086, BC-4032, BC-4199, HR3-61031, and HR6-527). Three of these BC samples provided the critical recombination events used to define the proximal (BC-4197 and BC-4086) and distal (BC-4199) boundaries for the Minimsc candidate interval reported previously (Hartmann et al. 2008). High-quality, high-molecular weight DNA was extracted using phenol–chloroform. DNA concentration was normalized to 50 ng/ml, processed according to the Affymetrix GenomeWide Human SNP Nsp/Sty Assay Kit 5.0/6.0 protocol, and hybridized to the Affymetrix MDA genotyping array (Yang et al. 2009) at the Functional Genomics Core Facility at the University of North Carolina at Chapel Hill (UNC) as previously described (Aylor et al. 2011). SNP genotype calling was conducted as described previously (Yang et al. 2009). The Minimsc candidate interval on MMU11 was refined using the

Figure 1 Flow diagram detailing each successive step, from phenotype to genotype, in the identification of Myh4Minimsc. (A) The mini-muscle was first observed in some of the HR and C lines (shown in red) but not in others (shown in black). The phenotype became fixed in the HR3 line. (B) A large backcross between HR3 and C57BL/6J was used to map the Minimsc locus. Two mice, BC-4197 and BC-4199, provided the initial boundaries for the Minimsc candidate interval. (C) High-density genotyping of these two samples provided more accurate boundaries for the candidate interval, including the proximal extension of the candidate interval. (D) Based on high-density genotyping, two HR samples with identical genotypes in the candidate interval but contrasting muscle phenotype were selected for whole-genome sequencing. (E) Identification of candidate genetic variants for the mini-muscle phenotype. (F) Discrimination between candidate variants and validation of Myh4 as the candidate gene.

heterozygous regions of mice BC-4197 and BC-4199 (BC mice delimiting the candidate region; see Hartmann et al. 2008) to fine-map boundaries of the candidate region (Figure 1). Hybridization intensities at every probe within the Minimsc candidate interval were compared between the minimuscle and normal mice in an effort to identify potential deletions and variable intensity oligonucleotides (Yang et al. 2011; Didion et al. 2012). Intensity plots spanning from 66.924 to 70.027 Mb on MMU11 among all pairwise comparisons of a mini-muscle and a normal sample were analyzed. Probes with differential hybridization were noted for each pair (mini-muscle and normal) and then compared across

the whole population. Mini-muscle, normal, and standard inbred strains with whole-genome sequence publicly available [referred as the Sanger strain set (Keane et al. 2011)] were compared using probe intensity plots and genotype calls to select the pair of samples with contrasting phenotypes (i.e., mini-muscle and normal) and most similar probe intensity plots for whole-genome sequencing and analysis. Whole-genome sequencing and alignment

Five micrograms of high-molecular-weight DNA from samples HR3-61031 and HR7-61023 were shipped from UNC to The Genome Institute at Washington University for

Novel SNP in Myh4 Reduces Hind-Limb Mass

1387

whole-genome sequencing. Genomic DNA was used to construct two Illumina HiSeq libraries, using 0.5 mg starting material, with 300- to 400- and 400- to 500-bp fragment sizes. Each library was sequenced on four lanes of an Illumina HiSeq flowcell, as paired-end reads, with 100-bp read lengths. This produced 155 Gb of high-quality sequence with 41.3083 haploid coverage for HR3-61031 and 120 Gb of high-quality sequence with 32.1883 haploid coverage for HR7-61023. We aligned the sequences to the University of California at Santa Cruz Mouse Build mm9 using ELAND to assess base accuracy; BAM files were then produced. We used Bowtie version 0.12.7 and aligned the newly generated sequence data to the mouse reference sequence [July 2007 (NCBI37/mm9)] under the following conditions: (1) Alignments may have no more than two mismatches in the first 28 bases on the high-quality (left) end of the read (the seed). (2) The sum of the Phred quality values at all mismatched positions may not exceed 70. Although we aligned sequence data to assembly [July 2007 (NCBI37/mm9)], all positions have been converted and are presented here to reflect the most recent assembly [December 2011 (GRCm38/ mm10)]. Only a portion of the whole-genome sequence was utilized to identify candidate mutations for Minimsc, and these sequences are provided for HR3-61031 and HR7-61023 in the Supporting Information (File S1 and File S2, respectively). File S1 and File S2 are provided in pileup format and are accompanied by a word document with an explanation of the SAMtools pileup format (File S3). Identification of candidate mutations for Minimsc

We used consistency and coverage as quality-control metrics to determine the usefulness of the sequence data. The two samples used in this analysis, HR3-61031 and HR7-61023, are homozygous in the Minimsc candidate interval and thus are expected to be homozygous at every position. We classified a given position as consistent when .65% of reads at that position had the same genotype call. Both samples were consistent at .99.5% of positions, and consistency was highly correlated between samples across the candidate interval. Similarly, coverage in each sample varied substantially across the candidate interval in both samples, but coverage was highly correlated between the two samples at a given position. To identify candidate SNPs, we used a custom script to find any position with more than a 10-read coverage in any of the two samples that had different genotype calls between the two samples. Genotyping at Ntn1 and Myh4 SNPs

To validate and extend the whole-genome sequence results, we genotyped a diverse set of samples using dideoxy Sanger sequencing of PCR amplicons spanning each one of the two candidate SNPs. PCR primers were designed to amplify 289and 323-bp-long PCR products in the Myh4 and Netrin 1 (Ntn1) genes, respectively. Primers were designed using Primer Quest software (http://www.idtdna.com). The primers for the Myh4 SNP are 59-AGGTGGCCATTTACAAGCTCACAG-39 and 59-GGGAATGGACAGAAACAGTAAAGGTCC-39 (MMU11:

1388

S. A. Kelly et al.

67,244,634–67,244,922). The primers for the Ntn1 SNP are 59-AGGATTTGCTGTCTGGCCCTCATT-39 and 59-ATGAAGTGCTCCTGGGTGCTTTCT-39 (MMU11: 68,282,072–68,282, 395). The PCR fragment was amplified using standard conditions (initial denaturation at 95° for 5 min, followed by 35 cycles of 95°, 55°, and 72° for 30 sec each, and a final extension at 72° for 10 min) in an Applied Biosystems Veriti 96 Well Thermal Cycler. After the PCR amplification, the product was purified using a Promega SV gel and Wizard Clean Up System. Sequencing of the purified PCR product was performed at the UNC Genome Analysis Core Facility using the Applied Biosystems 3730xl Genetic Analyzer. Sequences were then analyzed using Sequencher. Samples genotyped from various current and historical generations of the HR lines are summarized in Table 2. In addition, the following inbred strains were genotyped using the Myh4 and Ntn1 primers: BULS, BUSNA, CALB/RkJ, CZECHI/EiJ, CZECHII/EiJ, DDY/JclSidSeyFrkJ, JFI/Ms, KK/HlJ, LEWES/EiJ, LG/J, MOLD/DnJ, MOLF/EiJ, NON/ShiJ, NZW/ LacJ, PERC/EiJ, POHN, PWD/PhJ, RBF/DnJ, RIIIS/J, SEB, SEG, SFM, SKIVE/EiJ, SMON, STLT, STUF, WMP/PasDnJ, XBS, 129S1/SvImJ, AKR/J, A/J, BALB/cJ, C3H/HeJ, C57BL/ 6J, CAST/EiJ, CBA/J, DBA/2J, FVB/NJ, LP/J, NOD/ShiLtJ, NZO/HILtJ, PWK/PhJ, SPRET/EiJ, and WSB/EiJ. Finally, genotypes from the following lines were obtained from reported whole-genome sequencing (Keane et al. 2011): 129P2/OlaHsd, 129S1/SvImJ, 129S5SvEbBrd, AKR/J, A/J, BALB/cJ, C3H/ HeJ, C57BL/6NJ, CAST/EiJ, CBA/J, DBA/2J, LP/J, NOD/ ShiLtJ, NZO/HILtJ, PWK/PhJ, SPRET/EiJ and WSB/EiJ. The phylogenetic origin of the haplotype in the candidate region in these strains has been determined using MDA (Yang et al. 2011)

Results High-density genotyping refines the candidate interval for Minimsc

The MDA genotypes of samples BC-4197 and BC-4199 provide the new proximal and distal boundaries, respectively, for the Minimsc candidate interval (Figure 2). Specifically, the interval starts between positions 67,110,081 and 67,125, 771 [defined by MDA markers JAX00314160 and JAX00314161 (rs26906842), respectively] and ends between 70,209,794 and 70,210,630 [MDA markers JAX00028760 (rs29423020) and JAX00028761 (rs26921141), respectively]. The MDA genotype analysis refines the boundaries, but it also expands the previous candidate interval from the 2.6 Mb reported in Hartmann et al. (2008) to 3.1 Mb (Figure 2). This discrepancy is explained by a strong likelihood that the one microsatellite marker providing the original proximal boundary (D11Mit115) was yielding incorrect genotypes in BC-4197 and BC-4086 (compare the respective genotypes in Figure 2 with those reported in table 5 in Hartmann et al. 2008). Note also that this extended and better-annotated interval was critical for the identification of the causal variant (i.e., the causative SNP lies outside the initial candidate interval) (Figure 2 and see below).

Table 2 Genotyped samples from current and historical generations of replicated lines selectively bred for high voluntary wheel running Sample Line Generation Ntn1 SNPa Myh4 SNP Expected muscle phenotype Observed muscle phenotype Concordance for Myh4 64358 64426 64485 64382 64728 25087 25088 25091 25101 25126 25127 25243 25285 25289 25324 25330 25359 25360 25435 25440 25460 25465 25474 45295 45580 45581 45588 61031 64262 64301 64596 64170 64352 25001 25004 25120 25122 25182 25185 25279 25281 25293 25294 25319 25444 25447 25449 25537 25539 25574 25575 64406 64413 257427 25031 25038 25039 25040 25072 25075 25193 25196

1 1 1 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6

61 61 61 61 61 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 42 42 42 42 58 61 61 61 61 61 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 61 61 22 22 22 22 22 22 22 22 22

T T T T T ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND C C C C C C C C T T ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND T T ND ND ND ND ND ND ND ND ND

C C Het C C T T T T Het C Het T T Het T T T T C T T T T T T T T T T T C C C C Het Het Het C Het T C C C Het Het Het C C Het Het C C C Het C Het Het T Het T T

Normal Normal Normal Normal Normal Mini Mini Mini Mini Normal Normal Normal Mini Mini Normal Mini Mini Mini Mini Normal Mini Mini Mini Mini Mini Mini Mini Mini Mini Mini Mini Normal Normal Normal Normal Normal Normal Normal Normal Normal Mini Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Mini Normal Mini Mini

Normal Normal Normal Normal Normal Mini Mini Mini Mini Normal Normal Normal Mini Mini Normal Mini Mini Mini Mini Normal Mini Mini Mini Mini Mini Mini Mini Mini Mini Mini Mini Normal Normal Normal Normal Normal Normal Mini Normal Normal Mini Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Mini Normal Mini Mini

Yes Yes Yesb Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yesc Yesc Yesc Yesc NOd Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yesc Yese Yese Yese Yese Yese Yese Yese Yese (continued)

Novel SNP in Myh4 Reduces Hind-Limb Mass

1389

Table 2, continued Sample Line Generation Ntn1 SNPa Myh4 SNP Expected muscle phenotype Observed muscle phenotype Concordance for Myh4 25478 25483 25584 25586 25598 25601 25606 25615 25633 25636 25652 25658 40035 40036 40038 40040 40042 40044 41890 41892 41895 41896 41897 41899 64200 64201 64214 64222 64839 66015 66320 66326 66333 66346 66347 66348 66355 66357 66360 66367 66590 61023 64103 64135 64328

6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 8

22 22 22 22 22 22 22 22 22 22 22 22 37 37 37 37 37 37 38 38 38 38 38 38 61 61 61 61 61 63 63 63 63 63 63 63 63 63 63 63 63 58 61 61 61

ND ND ND ND ND ND ND ND ND ND ND ND T T T T T T T T T T T T T T T T T ND ND ND ND ND ND ND ND ND ND ND ND T T T T

Het Het Het Het C Het T Het T T T T C Het C T T T T T Het Het Het T Het T Het T Het T Het C Het Het Het Het T T T T T C C C C

Normal Normal Normal Normal Normal Normal Mini Normal Mini Mini Mini Mini Normal Normal Normal Mini Mini Mini Mini Mini Normal Normal Normal Mini Normal Mini Normal Mini Normal Mini Normal Normal Normal Normal Normal Normal Mini Mini Mini Mini Mini Normal Normal Normal Normal

Normal Normal Normal Normal Normal Normal Mini Normal Mini Mini Mini Mini Normal Normal Normal Mini Mini Mini Mini Mini Normal Normal Normal Mini Normal Mini Normal Mini Normal Mini Normal Normal Normal Normal Normal Normal Mini Mini Mini Mini Mini Normal Normal Normal Normal

Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yese Yes Yes Yes Yes

Samples were genotyped (C; T; Het, heterozygous) for Myh4 and Ntn1. The expected hind-limb muscle phenotype (based on genotype) was compared with the actual phenotype revealed post dissection. Concordance between the expected muscle phenotype based on the Myh4 genotype and the observed phenotype is presented in the far right column. a ND, sample not genotyped for respective SNP. b There is concordance between the Myh4 genotype and the observed phenotype. However, despite the presence of the variant in this sample, the mini-muscle phenotype has never been observed in line 1. c There is concordance between the Myh4 genotype and the observed phenotypes in all cases except one. Note that the mini-muscle phenotype was observed at least through generation 22 in line 5. d We hypothesize that there is an error in identification, recording, or labeling involving the animal, tissue sample, or extracted DNA from this discordant individual. Note that this sample originated from the historical samples dating back to generation 22 of selection in the year 1999. e Note that line 6 is still polymorphic for the mini-muscle phenotype as of generation 66.

High-density genotyping identifies the pair of optimal samples for identification of the candidate mutation Minimsc through whole-genome sequencing

The six samples genotyped with MDA from four HR selection line replicates are all homozygous within the Minimsc candidate interval (Figure 2). All samples from the HR7,

1390

S. A. Kelly et al.

HR3, and HR6 lines have identical haplotypes throughout the entire region (Figure 2, HR red haplotype), despite the fact that HR7 has the normal phenotype while HR3 and HR6 have the mini-muscle phenotype. These results confirm that Minimsc arose as a spontaneous mutation within the same genetic background shared by different HR lines. The HR8

line has a different haplotype (Figure 2, HR blue haplotype) starting between 69,818,782 and 69,818,849 (JAX00314705 and JAX00314706, respectively) and continuing through the end of the Minimsc candidate interval. This change in haplotype covering the distal 392 kb of the Minimsc candidate region is unrelated to the mini-muscle phenotype. Analysis of SNP and exon probe intensity plots shows no evidence of large deletions or insertions in the region (results not shown). Genotype calls and intensity plots were also used to determine which pair of HR samples showed the strongest evidence for identity by decent (IBD) within the Minimsc candidate interval between samples with the mini-muscle and normal phenotypes. This analysis indicated that the closest match was between sample HR3-61031 with the mini-muscle phenotype and sample HR7-61023 with the normal phenotype (Figure 2). Only three probes had significantly different probe intensities between these two samples. Other pairwise combinations involving combinations of HR samples with discordant mini-muscle phenotypes have additional probes with significantly different intensities. Therefore, we concluded that HR3-61031 and HR7-61023 represented the best pair of samples that were genotypically matched but phenotypically discordant. Thus, given the available choices, these two samples are optimal for directly identifying candidate mutations for Minimsc as any genetic variant that distinguishes them within the new boundaries of the candidate interval using whole-genome sequencing. We also conclude that among the set of Sanger inbred strains (Keane et al. 2011) the CBA/J strain shows the strongest evidence for IBD in the Minimsc candidate region with HR3-61031 based on MDA genotypes (data not shown). Whole-genome sequence identifies two candidate mutations for Minimsc

We used the aligned reads from the two sequenced samples (HR3-61031 and HR7-61023) to identify candidate genetic variants for Minimsc. We first determined the coverage over the candidate and flanking regions (at least 100 kb proximal and distal flanking regions; Figure S1). We used these data to determine the threshold used in variant calling (see below). We also determined whether there was differential coverage between these two samples in any 500-bp windows. We found strong evidence of a structural variant at the Nlrp1 gene family locus (71.09–71.29 Mb). This result demonstrates our ability to detect copy-number variations. However, this deletion can be ruled out as underlying Minimsc because it is shared by the two samples with discordant phenotypes and it is located outside of the candidate interval. In the 3.1-Mb Minimsc refined candidate interval, we identified only two SNPs with different alleles in our two samples. These two SNPs are located at positions 67,244, 850 and 68,282,254. The first SNP is a C-to-T transition located in a 709-bp intron between exons 11 and 12 of the Myh4 skeletal muscle gene. The second SNP is a T-to-C transition in a very long intron (.100 kb) separating exons 3 and 4 of the Ntn1 gene. The variant allele at both SNPs is found in

the HR3-61031 sample that has the mini-muscle phenotype [i.e., this allele is not found in the mouse reference genome nor in any inbred strain genotyped or sequenced at that position (Table S1)], and it is well supported by forward and reverse reads with a good read coverage (minimum coverage is 24 reads). Because it was not immediately obvious whether either of these SNPs would have functional consequences, we explored the possibility of additional undiscovered SNPs in regions with low coverage or at positions with inconsistent genotype calls within a sample. We could reject both scenarios with a high degree of confidence by comparing coverage and the ratio of inconsistent genotypes between the two sequenced samples. Coverage and inconsistent SNPs were almost identical at both samples. Additionally, there was no evidence of any insertion in the HR3-61031 sample based on disruptions of forward and reverse read density (data not shown). We concluded that one of the two SNPs in either the Myh4 or the Ntn1 genes was the cause of the mini-muscle phenotype and thus represents Minimsc. Extensive genotyping in additional samples excludes the Ntn1 SNP and provides independent evidence supporting the Myh4 SNP as Minimsc

The variant allele at both the Ntn1 and Myh4 SNPs is not present among 46 inbred strains genotyped by us or for which genome sequence had been reported recently (Table S1; Keane et al. 2011). These strains represent a wide variety of classical and wild-derived stocks from multiple Mus taxa including M. musculus musculus, M. m. domesticus, M. m. castaneus, M. spretus, and M. macedonicus. These results indicate that the alleles identified in the HR stocks at both SNPs are derived alleles of recent origin. We genotyped both SNPs in additional HR samples that had been phenotyped for presence or absence of the minimuscle. Table 2 summarizes the results including the concordance between genotype and phenotype. For the Ntn1 SNP there are eight discordant samples; all of them have the minimuscle phenotype and are homozyogous for the reference allele. This level of discordance and the fact that only HR3 samples have the alternative allele, rules out this SNP as a candidate for Minimsc. In contrast, all but one sample of 107 genotyped at the Myh4 SNP are concordant with the minimuscle phenotype (Table 2), including samples from both HR lines that express the mini-muscle phenotype (HR3 and HR6) and historical samples dating from August 1999 (samples beginning with the number 25). The single discordant sample (25182) is heterozygous at the Myh4 SNP and reportedly has the mini-muscle phenotype. We hypothesize that there is an error in identification, recording, or labeling involving the animal, tissue sample, or extracted DNA from this discordant individual. This is quite plausible given that this sample originated from the historical samples dating back to generation 22 of selection in the year 1999. We therefore conclude that the Myh4 SNP is Minimsc (allele symbol, Myh4Minimsc) and that the Ntn1 variant represents a more recent SNP that arose and was fixed in the HR3 selection line.

Novel SNP in Myh4 Reduces Hind-Limb Mass

1391

Figure 2 High-density genotyping refines the candidate interval for Minimsc. The different haplotypes present in each chromosome of each sample genotyped are shown as lines of different colors. Squares represent actual genotypes (MDA, microsatellites, and PCR sequencing). Red represents genotypes and haplotypes present in HR lines. Blue denotes alternative genotypes and haplotypes present in a subset of HR lines. Black represents genotypes and haplotypes present in C57BL/6J inbred line.

Discussion Transcriptional profiling of muscle shows a significant reduction in expression of Myh4 in Myh4Minimsc relative to wild type

Previously, genome-wide expression profiling described transcriptome differences in the medial gastrocnemius between HR mice (generation 37) with the mini-muscle phenotype (line 3) and HR mice with wild-type muscles (lines 7 and 8) (Burniston et al. 2013). Differential expression analysis revealed that, consistent with the known reduction of type IIB fibers (Guderley et al. 2008), Myh4 expression was expressed at significantly lower levels (28.82-fold reduction) in muscles taken from adult mice with the mini-muscle phenotype (Burniston et al. 2013). This change was accompanied by additional expression differences in structural genes [e.g., myocyte enhancer factor 2C (Mef2c), 2-fold greater in mini-muscle samples] and myogenic factors [e.g., myogenin (Myog), 3.8-fold greater in mini-muscle] associated with slow-type myofibrils. These findings are an important functional validation of the results presented here and further support that Myh4Minimsc is responsible for the mini-muscle phenotype. Phenotypic consequences of the mini-muscle mutation are strikingly similar in many ways to those observed in a strain of mice lacking IIB myosin heavy chain due to targeted disruption of the Myh4 gene, which was first reported by Acakpo-Satchivi et al. (1997) and later characterized by Allen et al. (2001). In general, this includes changes in muscle mass, reduced body mass, larger hearts, no overt differences in locomotor behavior, reduced grip strength (K. M. Middleton, R. M. Hannon, and T. Garland Jr., unpublished results), reduced muscle force (Syme et al. 2005; McGillivray et al. 2009), and ultrastructural and histological differences (Guderley et al. 2006, 2008). In contrast, several findings in the Myh4 knockout strain do not closely match data from mini-muscle mice. For example, mini-muscle mice have longer

1392

S. A. Kelly et al.

tibiafibulae (Kelly et al. 2006), eat more food per gram of body mass (Meek et al. 2010), have disorganized muscle fibers in certain muscle regions (Guderley et al. 2006, 2008), and have reduced contractile speed (Syme et al. 2005; McGillivray et al. 2009), whereas none of those alterations have been reported in the Myh4 knockout mice. Some of the differences between the mice may be attributable to context (specific muscle studied, age, sex, rearing environment), while others may be due to the different types of mutations (null in the knockout and likely hypomorphic in the minimuscle mice. In addition, they could also be the consequence of genetic background, which can exert strong influences on expressions of genes of major effect. As examples, the minimuscle mutation shows statistically different effects on body mass, muscle mass, relative composition of myosin heavy and light chains, and mitochondrial volume densities per plantaris muscle fiber in the two HR lines that exhibit the phenotype (Guderley et al. 2006). With regard to voluntary wheel running, Harrison et al. (2002) reported that myosin heavy chain null mice (IIb and IId/x) had reduced running performance [both distance (km/night) and average speed (m/min)] compared to nontransgenic C57BL/6. Although the myosin heavy chain null mice ran slower and for shorter distances compared to nontransgenics, muscles of both null groups showed an ability to train (skeletal muscle plasticity) or proximately adapt to exercise (Harrison et al. 2002). In response to voluntary wheel running, both null groups showed increases in muscle fiber size and muscle oxidative capacity (Harrison et al. 2002). In comparison, mini-muscle individuals also show an ability to train in response to voluntary wheel running, as demonstrated by Houle-Leroy et al. (2000, 2003). However, it is important to note that, unlike the myosin heavy chain null mice discussed above, mini-muscle individuals do not consistently run less on wheels (revolutions/day) than other HR mice (Garland et al. 2002).

Potential biomedical significance of Myh4Minimsc

Variable muscle function and physiological properties underlying muscular performance are in part attributable to different isoforms of the molecular myosin heavy chain. Heritable mutations in the various isoforms have been associated with muscular syndromes and diseases. The pathology and mechanisms of these myosin myopathies have been detailed elsewhere (Tajsharghi and Oldfors 2013). Phenotypically, myosin myopathies broadly display disintegration of the myofibrillar network and accumulation of cellular protein aggregates. Collectively, when grouped with myofibrillar myopathies (also displaying protein aggregates), these myopathies are termed “protein aggregate myopathies” (Goebel and Muller 2006). The underlying genetic architecture and the resulting molecular pathway leading to phenotypic consequences is still undetermined for many of these myopathies, although protein conformational changes as a result of mutations seem a likely molecular cause (Ferrer and Olive 2008; Schroder and Schoser 2009). Myofibrillar myopathy has been observed in a mouse mutant model (ariel) as a result of an ENU-induced recessive mutation resulting in a change (L342Q) within the motor domain of Myh4 (MyHC IIb) (Kurapati et al. 2012). By day 13 (concomitant with the switch from developmental to adult myosin isoforms), homozygotes (Myh4arl/Myh4arl) develop limb paralysis caused by the aggregate-prone protein (Kurapati et al. 2012). No other tissue displayed pathological changes (Kurapati et al. 2012). The phenotypes (e.g., muscle atrophy, rigid gait, muscle paralysis) in this ariel mutant are in sharp contrast to the Myh4 null mice discussed above and the Myh4Minimsc mice discussed here. The differences between Myh4 null and Myh4arl/Myh4arl mice, as hypothesized by Kurapati et al. (2012), may be in part attributable to compensatory mechanisms (muscle fiber hypertrophy, upregulation of the adult myosin heavy chain IIa and IId) chronicled in Myh4 null mice (Allen et al. 2001). Muscle fiber hypertrophy and upregulation of the adult myosin heavy chain IIa have previously been chronicled in Myh4Minimsc individuals (Guderley et al. 2008). We also hypothesize that phenotypic differences between Myh4Minimsc and Myh4arl/Myh4arl mice are caused by similar compensatory mechanisms as observed in Myh4 null mice. Regardless, all three models may serve to elucidate the downstream molecular mechanisms leading to drastically different phenotypic outcomes as a result of variable changes in Myh4. In an attempt to determine if the Myh4Minimsc variant is unique to the current population, or potentially appears in other species, we utilized a standard nucleotide BLAST (BLASTn) to query 115 bp before and after Myh4Minimsc. BLASTn returned zero matching sequences inclusive of the Myh4Minimsc variant from nonmurine species. We agree with Kurapati et al. (2012) that establishing genotype–phenotype correlations for myosin mutations is a challenging exercise, but it has now been accomplished in at least three different mouse models for mutated Myh4.

Although Myh4 (MyHC IIb) is an essential motor protein in rats and mice, in humans the messenger RNA has been found only in masseter muscles (e.g., Horton et al. 2001). However, functional equivalents have been identified in humans (Myh1), and a patient with a mutation in Myh7 had a phenotypically similar myopathy to the Myh4arl/ Myh4arl mouse (Tasca et al. 2012). Moreover, Harrison et al. (2011) observed that MyHC IIb is expressed in fetal human cells in vitro and may be “reactivated” in degenerative/regenerative adult muscles, thus strengthening the potential for translational relevance of the work presented here. Evolutionary significance of Myh4Minimsc

As noted previously (Hannon et al. 2008), the Myh4Minimsc allele exhibits classical properties of a gene of major effect (GOME), including its Mendelian recessive nature, dramatic effect on phenotype (muscle mass), and numerous pleiotropic effects. Although the modern synthesis generally emphasized the importance of alleles with relatively small effects for phenotypic evolution in the wild, the potential importance of GOMEs has long been appreciated in the world of artificial selection and domestication, and the potential importance of GOMEs in the wild has also been emphasized in recent years (Stinchcombe et al. 2009). Theoretical studies have shown that GOMEs can alter genetic architecture dramatically for brief periods of time and hence change the rate and direction of multivariate evolution in important ways (Agrawal et al. 2001). Various empirical studies have also suggested the importance of GOMEs for evolution in the wild (e.g., Voss and Shaffer 1997; Voss and Smith 2005). The mini-muscle phenotype, the underlying Myh4Minimsc mutation discussed here, and its potential evolutionary significance are fairly unique in the context of experimental evolution (Garland and Rose 2009). GOMEs as a product of experimental evolution are typically detected after fixation, inhibiting the examination of their effect on selection, variation, or covariation of traits (Lande 1979). In the current model, we have the ability to “watch” this mutation in both a historical and a current context, as it has disappeared in one control line, gone to fixation in one selection line, and remains polymorphic in an additional selection line. Although other examples exist (see Stinchcombe et al. 2009), we feel the discovery and subsequent identification of this mutation will provide an opportunity to more thoroughly examine the role of GOMEs in the response to selection. Another benefit of identifying a GOME in the context of this particular selection experiment is that we have a plethora of data on traits other than the selected phenotype (voluntary wheel running) (see references in Garland et al. 2011b; Burniston et al. 2013). This will enable us to better understand how the mutation has affected and might affect variation in related suborganismal traits related to voluntary wheel running. Significant interactions have previously been identified between Myh4Minimsc and loci underlying voluntary wheel running and tissue mass traits (Nehrenberg et al.

Novel SNP in Myh4 Reduces Hind-Limb Mass

1393

2010). Future work will build upon these results and continue to determine how Myh4Minimsc interacts with additional minor-effect loci to determine phenotypes observed in individuals with the mini-muscle phenotype (e.g., Papa et al. 2013). Additionally, it will be a priority to determine what genetic mechanism is facilitating these interactions (pleiotropy, physical linkage/linkage disequilibrium, or epistasis). Methodological significance of narrowing a broad locus to two candidate SNPs

Here we combined the benefits of experimental evolution with the tools of modern genomics to discover an adaptive phenotype inherited in a Mendelian recessive nature; to map the underlying genetic architecture to a 3.1-Mb interval on MMU11, a region containing 100 expressed or predicted genes; and to identify the causal SNP responsible for the mini-muscle phenotype. Typically, identification of the causative variant responsible for a given phenotype is limited by the presence of extensive linkage disequilibrium in both human and animal models and by our incomplete knowledge of how genetic variants affect the function of nearby genes. However, it is possible to identify the causative variant if these mutations are recent and within within a known haplotype. Here we combined standard linkage mapping with the identification and deep sequencing of two samples with discordant phenotypes with identical haplotypes in the candidate region to identify two candidate SNPs in a 3.1-Mb interval. Finally, we were able to exclude these two variants by performing a simple association analysis in a relatively small set of additional samples. Using next-generation sequencing, this approach can be extended to identify causative mutations arising in mouse inbred lines and thus offers a great avenue to overcome one of the most challenging steps in quantitative genetics. Note that exome sequencing, a standard approach to identifying GOMEs (e.g., Ng et al. 2010), would have failed to identify the causative SNP. Despite the success of our approach, the mechanism by which an intronic SNP affects the function of the Myh4 gene remains unclear. The location of the SNP, the recessive nature of the allele, and the expression data strongly suggest a regulatory variant. However, future investigations will seek to determine if there is an association between the variant and lower levels of protein expression, which would lend further support to the regulatory variant hypothesis. Additionally, validation will require generation and analysis of mice with specific genotype combinations at Myh4 during embryonic and early postnatal development.

Acknowledgments We thank T. Meek and V. Careau for assistance in obtaining muscle samples and the DNA sequencing production team led by Lucinda Fulton and Robert Fulton at the Genome Institute at Washington University for sample management and data production. This work was partially supported by

1394

S. A. Kelly et al.

National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) grant DK076050 (to D.P.). Some phenotypes were collected using the Animal Metabolism Phenotyping core facility within the University of North Carolina’s Nutrition Obesity Research Center (funded by NIDDK grant DK056350). Additional support was provided by National Science Foundation grants IOS-1121273 and IOB-0543429 (to T.G.). Support for the development of genomic tools and pipelines for array and whole-genome sequence analysis was provided by National Institute of Mental Health/National Human Genome Research Institute Center of Excellence for Genome Sciences grants P50MH090338 and P50HG006582 (to F.P.-M.d.V.).

Literature Cited Acakpo-Satchivi, L. J. R., W. Edelmann, C. Sartorius, B. D. Lu, P. A. Wahr et al., 1997 Growth and muscle defects in mice lacking adult myosin heavy chain genes. J. Cell Biol. 139: 1219–1229. Agrawal, A. F., E. D. Brodie, and L. H. Rieseberg, 2001 Possible consequences of genes of major effect: transient changes in the g-matrix. Genetica 112–113: 33–43. Allen, D. L., B. C. Harrison, C. Sartorius, W. C. Byrnes, and L. A. Leinwand, 2001 Mutation of the IIB myosin heavy chain gene results in muscle fiber loss and compensatory hypertrophy. Am. J. Physiol. Cell Physiol. 280: C637–C645. Aylor, D. L., W. Valdar, W. Foulds-Mathes, R. J. Buus, R. A. Verdugo et al., 2011 Genetic analysis of complex traits in the emerging Collaborative Cross. Genome Res. 21: 1213–1222. Burniston, J. G., T. H. Meek, S. N. Pandey, G. Broitman-Maduro, M. F. Maduro et al., 2013 Gene expression profiling of gastrocnemius of ‘Mini-Muscle’ mice. Physiol. Genomics 45: 228–236. Careau, V. C., O. R. P. Bininda-Emonds, G. Ordonez, and T. Garland, Jr., 2012 Are voluntary wheel running and open-field behavior correlated in mice? Different answers from comparative and artificial selection approaches. Behav. Genet. 42: 830–844. Careau, V., M. E. Wolak, P. A. Carter, and T. Garland, Jr., 2013 Limits to behavioral evolution: the quantitative genetics of a complex trait under directional selection. Evolution DOI: 10.1111/evo.12200. Carriére, Y., J. P. Deland, D. A. Roff, and C. Vincent, 1994 Life history costs associated with the evolution of insecticide resistance. Proc. Biol. Sci. 258: 35–40. Didion, J. P., H. Yang, K. Sheppard, C. P. Fu, L. McMillan et al., 2012 Discovery of novel variant in genotyping arrays improves genotype retention and reduces ascertainment bias. BMC Genomics 13: 34. Ferrer, I., and M. Olive, 2008 Molecular pathology of myofibrillar myopathies. Expert Rev. Mol. Med. 10: e25. Garland, T. Jr., and M. R. Rose, 2009 Experimental Evolution: Concepts, Methods, and Applications of Selection Experiments. University of California Press, Berkeley, CA. Garland, T. Jr., M. T. Morgan, J. G. Swallow, J. S. Rhodes, I. Girard et al., 2002 Evolution of a small-muscle polymorphism in lines of house mice selected for high activity levels. Evolution 56: 1267–1275. Garland, T. Jr., S. A. Kelly, J. L. Malisch, E. M. Kolb, R. M. Hannon et al., 2011a How to run far: multiple solutions and sex-specific responses to selective breeding for high voluntary activity levels. Proc. Biol. Sci. 278: 574–581. Garland, T. Jr., H. Schutz, M. A. Chappell, B. K. Keeney, T. H. Meek et al., 2011b The biological control of voluntary exercise,

spontaneous physical activity and daily energy expenditure in relation to obesity: human and rodent perspectives. J. Exp. Biol. 214: 206–229. Girard, I., J. G. Swallow, P. A. Carter, P. Koteja, J. S. Rhodes et al., 2002 Maternal-care behavior and life-history traits in house mice (Mus domesticus) artificially selected for high voluntary wheel-running activity. Behav. Processes 57: 37–50. Goebel, H. H., and H. D. Muller, 2006 Protein aggregate myopathies. Semin. Pediatr. Neurol. 13: 96–103. Guderley, H., P. Houle-Leroy, G. M. Diffee, D. M. Camp, and T. Garland, Jr., 2006 Morphometry, ultrastructure, myosin isoforms, and metabolic capacities of the “mighty mini muscles” favoured by selection for high activity in house mice. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 144: 271–282. Guderley, H., D. R. Joanisse, S. Mokas, G. M. Bilodeau, and T. Garland, Jr., 2008 Altered fiber types in gastrocnemius muscle of high wheel-running selected mice with mini muscle phenotypes. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 149: 490– 500. Hannon, R. M., S. A. Kelly, K. M. Middleton, E. M. Kolb, D. Pomp et al., 2008 Phenotypic effects of the “mini-muscle” allele in a large HR 3 C57BL/6J mouse backcross. J. Hered. 99: 349– 354. Harrison, B. C., M. L. Bell, D. l. Allen, W. C. Byrnes, and L. A. Leinwand, 2002 Skeletal muscle adaptations in response to voluntary wheel running in myosin heavy chain null mice. J. Appl. Physiol. 92: 313–322. Harrison, B. C., D. L. Allen, and L. A. Leinwand, 2011 IIB or not IIB? Regulation of myosin heavy chain gene expression in mice and men. Skelet. Muscle 1: 5. Hartmann, J., T. Garland, Jr., R. M. Hannon, S. A. Kelly, G. Muñoz et al., 2008 Fine mapping of “Mini-Muscle,” a recessive mutation causing reduced hind-limb muscle mass in mice. J. Hered. 99: 679–687. Horton, M. J., C. A. Brandon, T. J. Morris, T. W. Braun, K. M. Yaw et al., 2001 Abundant expression of myosin heavy-chain IIB RNA in a subset of human masseter muscle fibres. Arch. Oral Biol. 46: 1039–1050. Houle-Leroy, P., T. Garland, Jr., J. G. Swallow, and H. Guderley, 2000 Effects of voluntary activity and genetic selection on muscle metabolic capacities in house mice Mus domesticus. J. Appl. Physiol. 89: 1608–1616. Houle-Leroy, P., T. Garland, Jr., J. G. Swallow, and H. Guderley, 2003 Artificial selection for high activity favors mighty minimuscles in house mice. Am. J. Physiol. Regul. Integr. Comp. Physiol. 284: R433–R443. Keane, T. M., L. Goodstadt, P. Danecek, M. A. White, K. Wong et al., 2011 Mouse genomic variation and its effect on phenotypes and gene regulation. Nature 477: 289–294. Kelly, S. A., P. P. Czech, J. T. Wight, K. M. Blank, and T. Garland, Jr.., 2006 Experimental evolution and phenotypic plasticity of hindlimb bones in high-activity house mice. J. Morphol. 267: 360–374. Kolb, E. M., S. A. Kelly, K. M. Middleton, L. S. Sermsakdi, M. A. Chappell et al., 2010 Erythropoietin elevates VO2,max but not voluntary wheel running in mice. J. Exp. Biol. 213: 510–519. Kurapati, R., C. McKenna, J. Lindqvist, D. Williams, M. Simon et al., 2012 Myofibrillar myopathy caused by a mutation in the motor domain of mouse MyHC IIb. Hum. Mol. Genet. 21: 1706– 1724. Lande, R., 1979 Quantitative genetic analysis of multivariate evolution, applied to brain: body size allometry. Evolution 33: 402– 416. McGillivray, D. G., T. Garland, Jr., E. M. Dlugosz, M. A. Chappell, and D. A. Syme, 2009 Changes in efficiency and myosin ex-

pression in the small-muscle phenotype of mice selectively bred for high voluntary running activity. J. Exp. Biol. 212: 977–985. Meek, T. H., J. C. Eisenmann, and T. Garland, Jr., 2010 Western diet increases wheel running in mice selectively bred for high voluntary wheel running. Int. J. Obes. 34: 960–969. Middleton, K. M., S. A. Kelly, and T. Garland, Jr., 2008 Selective breeding as a tool to probe skeletal response to high voluntary locomotor activity in mice. Integr. Comp. Biol. 48: 394–410. Nehrenberg, D. L., S. Wang, R. M. Hannon, T. Garland, Jr., and D. Pomp, 2010 QTL underlying voluntary exercise in mice: interactions with the “mini muscle” locus and sex. J. Hered. 101: 42– 53. Ng, S. B., K. J. Buckingham, C. Lee, A. W. Bigham, H. K. Tabor et al., 2010 Exome sequencing identifies the cause of a Mendelian disorder. Nat. Genet. 42: 30–35. Papa, R., D. D. Kapan, B. A. Counterman, K. Maldonado, D. P. Lindstrom et al., 2013 Multi-alleic major effect genes interact with minor effect QTLs to control adaptive color pattern variation in Heliconius erato. PLoS ONE 8: e57033. Rivero, A., A. Magaud, A. Nicot, and J. Veziliér, 2011 Energetic cost of insecticide resistance in Culex pipiens mosquitoes. J. Med. Entomol. 48: 694–700. Schroder, R., and B. Schoser, 2009 Myofibrillar myopathies: a clinical and myopathological guide. Brain Pathol. 19: 483–492. Stinchcombe, J. R., C. Weinig, K. D. Heath, M. T. Brock, and J. Schmitt, 2009 Polymorphic genes of major effect: consequences for variation, selection and evolution in Arabidopsis thaliana. Genetics 182: 911–922. Swallow, J. G., P. A. Carter, and T. Garland, Jr., 1998 Artificial selection for increased wheel-running behavior in house mice. Behav. Genet. 28: 227–237. Swallow, J. G., J. S. Rhodes, and T. Garland, Jr., 2005 Phenotypic and evolutionary plasticity of organ masses in response to voluntary exercise in house mice. Integr. Comp. Biol. 45: 426–437. Swallow, J. G., J. P. Hayes, P. Koteja, and T. Garland, Jr., 2009 Selection experiments and experimental evolution of performance and physiology, pp. 301–333 in Experimental Evolution: Concepts, Methods, and Applications of Selection Experiments, edited by T. Garland Jr., and M. R. Rose. University of California Press, Berkeley, CA. Syme, D. A., K. Evashuk, B. Grintuch, E. L. Rezende, and T. Garland, Jr., 2005 Contractile abilities of normal and “mini” triceps surae muscles from mice (Mus domesticus) selectively breed for high voluntary wheel running. J. Appl. Physiol. 99: 1308–1316. Tajsharghi, H., and A. Oldfors, 2013 Myosinopathies: pathology and mechanisms. Acta Neuropathol. 125: 3–18. Tasca, G., E. Ricci, S. Penttila, M. Monforte, V. Giglio et al., 2012 New phenotype and pathology features in MYH7-related distal myopathy. Neuromuscul. Disord. 22: 640–647. Voss, S. R., and H. B. Shaffer, 1997 Adaptive evolution via a major gene effect: paedomorphosis in the Mexican axolotl. Proc. Natl. Acad. Sci. USA 94: 14185–14189. Voss, S. R., and J. J. Smith, 2005 Evolution of salamander life cycles: a major-effect quantitative trait locus contributes to discrete and continuous variation for metamorphic timing. Genetics 170: 245–281. Yang, H., Y. Ding, L. N. Hutchins, J. Szatkiewicz, T. A. Bell et al., 2009 A customized and versatile high-density genotyping array for the mouse. Nat. Methods 6: 663–666. Yang, H., J. R. Wang, J. P. Didion, R. J. Buus, T. A. Bell et al., 2011 Subspecific origin and haplotype diversity in the laboratory mouse. Nat. Genet. 43: 648–655. Communicating editor: D. W. Threadgill

Novel SNP in Myh4 Reduces Hind-Limb Mass

1395

GENETICS Supporting Information http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.113.154476/-/DC1

A Novel Intronic Single Nucleotide Polymorphism in the Myosin heavy polypeptide 4 Gene Is Responsible for the Mini-Muscle Phenotype Characterized by Major Reduction in Hind-Limb Muscle Mass in Mice Scott A. Kelly, Timothy A. Bell, Sara R. Selitsky, Ryan J. Buus, Kunjie Hua, George M. Weinstock, Theodore Garland Jr., Fernando Pardo-Manuel de Villena, and Daniel Pomp

Copyright © 2013 by the Genetics Society of America DOI: 10.1534/genetics.113.154476

Figure S1 Coverage and sequence consistency in the two HR samples for which whole-genome sequence was generated in this study. (A) and (B) show the frequency distribution in pileup high in the updated candidate interval ± 1 Mb proximal and distal. (C) shows the coverage difference between the two samples along the entire interval calculated as: (HR3-HR7) - average (HR3-HR7). (D) shows the correlation between the number of consistent calls in the two samples in 200 bp windows across the candidate interval ± 1 Mb proximal and distal.

2 SI

S. A. Kelly et al.

Files S1-S2 Available for download at http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.113.154476/-/DC1

File S1 Whole-genome sequence of HR3-61031 utilized to identify candidate mutations for Minimsc

File S2 Whole-genome sequence of HR7-61023 utilized to identify candidate mutations for Minimsc

S. A. Kelly et al.

3 SI

File S3 SAMtools pileup format

Example: chr11 66823583 chr11 66823584 chr11 66823585 chr11 66823586 chr11 66823587 chr11 66823588 chr11 66823589 chr11 66823590 chr11 66823591

t t c c t c c c c

17 17 17 17 17 17 17 17 19

.c...,.,,..,.,.., .,...,.,,..,.,.., .,...,.,,..,.,.., .,...,.,,..,.,.., .,...,.,,..,.,.., .,...,.,,..,.,.., .,...,.,,..,.,.., .,...,.,,..,.,.., .,...,.,,..,.,..,^~.^~.

H-H9HH5CAHEEHBFH9 H(HEH@A;